%0 Journal Article %J Health Services Research %D 2021 %T Eligible but Not Enrolled: Consequences of Forgoing Prescription Drug Subsidies Among Low-Income Medicare Beneficiaries with Diabetes %A Glynn, Alexandra %A Hernandez, Inmaculada %A Roberts, Eric %K Diabetes %K low-income %K Medicare Beneficiaries %K prescription drugs %K subsidies %X Research Objective To estimate take-up of the Part D Low-Income Subsidy (LIS) among Medicare beneficiaries with diabetes, and to examine differences in out-of-pocket costs and prescription drug use between enrollees and LIS-eligible non-enrollees. The LIS lowers prescription drug costs for Medicare beneficiaries with incomes ≤150% of the Federal Poverty Level (FPL), and caps out-of-pocket costs for beneficiaries with incomes ≤135% of FPL. Although a number of recent policy proposals seek to address rising drug costs, the LIS itself remains underutilized. The value of forgoing this benefit among chronically ill beneficiaries has not previously been studied. Study Design We used data from the Health and Retirement Study (HRS) to estimate take-up of the LIS among Medicare beneficiaries with diabetes. Using propensity score-weighted analyses, we compared out-of-pocket costs, prescription drug use, and cost-related drug non-adherence among LIS enrollees and eligible non-enrollees. We stratified the analyses according to whether respondents' income was ≤100% vs. >100% to ≤150% of FPL because of differences in the generosity and take-up of Medicaid benefits above vs. below 100% of FPL, and because individuals with Medicaid automatically receive the LIS. Population Studied Community-dwelling HRS respondents surveyed biennially between 2008–2016 who were enrolled in Medicare Part D, had an established diagnosis of diabetes, and were eligible for the LIS (income ≤150% FPL and assets below LIS limit). Principal Findings Among Medicare beneficiaries with diabetes, 56.9% of those with incomes >100% to ≤150% of FPL received the LIS, while 90.3% of those with incomes ≤100% received the LIS, reflecting automatic enrollment from Medicaid. Based on these take-up rates, we estimate that 322,746 LIS-eligible individuals with diabetes do not enroll in the LIS annually. For those with incomes >100% to ≤150% of FPL, compared to LIS enrollees, eligible non-enrollees incurred higher annual out-of-pocket drug spending (\$518, p < 0.001), had 20% fewer fills for diabetes, hypertension, and hyperlipidemia prescription drugs (p = 0.003), and were 9 percentage points more likely to report skipping drugs due to cost (p = 0.04). Conclusions A substantial proportion of Medicare beneficiaries with diabetes are eligible for but not enrolled in the LIS. Low take-up was concentrated among beneficiaries with income >100% to ≤150% of FPL. Forgoing the LIS was associated with higher out-of-pocket costs and cost-related non-adherence, and lower use of medications used to treat chronic conditions. Our findings demonstrate the extent to which out-of-pocket costs may be reduced, and use of chronic disease medications increased, if a higher proportion of LIS-eligible Medicare beneficiaries receive this benefit. Implications for Policy or Practice Proposals to lower out-of-pocket drug costs for Medicare beneficiaries often overlook the fact that existing prescription drug subsidy programs are underutilized. Because the LIS reduces cost sharing for all Part D covered drugs and caps annual out-of-pocket costs for most recipients, increasing LIS take-up would likely provide better financial protection for seniors with chronic conditions than a flat subsidy or price caps limited to certain drugs. Policies to streamline or automate LIS enrollment for low-income Medicare beneficiaries could help to mitigate the impact of rising drug costs in this vulnerable population. Primary Funding Source Agency for Healthcare Research and Quality. %B Health Services Research %V 56 %P 62 %G eng %N S2 %R 10.1111/1475-6773.13784 %0 Journal Article %J Innovation in Aging %D 2020 %T Effects of Medicare Drug Subsidies on Adherence for Diabetics: Evidence From a Regression Discontinuity Design %A Glynn, Alexandra %A Hernandez, Inmaculada %A Roberts, Eric %K diabetics %K medicare drug subsidies %X Out-of-pocket prescription drug costs are rapidly rising, particularly for insulin, which is a life-saving drug used by 3.1 million diabetics on Medicare. High out-of-pocket costs place an accentuated financial strain on older adults with diabetes, many of whom have low incomes, and may impede medication adherence, leading to poor health outcomes. The Medicare Part D Low-Income Subsidy (LIS) program limits drug co-pays to under $8.50 per prescription and caps out-of-pocket drug costs for lowest-income recipients (<135% Federal Poverty Level, FPL), resulting in pronounced differences in out-of-pocket costs for those with marginally different incomes. Using detailed income data from the Health and Retirement Study linked to Medicare claims (2008-2016), we employed a regression discontinuity (RD) design to isolate the effects of differences in out-of-pocket costs at eligibility thresholds for the LIS. Diabetic beneficiaries whose income exceeded the LIS eligibility threshold had lower Part D spending (-$945/year, p=0.03, n=2,367) and adherence to oral antidiabetic drugs (-8%, p=0.02). We conducted secondary analyses at the eligibility threshold for Medicaid, as individuals whose income exceeds the eligibility limit for Medicaid (100% of FPL in most states) are significantly less likely to receive the LIS. Above the Medicaid eligibility threshold (n=2,295), annual spending on insulin was $395 lower (p=0.002) and proportion of insulin use was 6% lower (p=0.04). These results suggest low-income Medicare beneficiaries who are not shielded from out-of-pocket costs via the LIS are particularly sensitive to drug costs. Policy proposals to limit out-of-pocket costs could improve medication adherence to high-cost drugs for vulnerable beneficiaries. %B Innovation in Aging %V 4 %P 280 %@ 2399-5300 %G eng %N Suppl 1 %R 10.1093/geroni/igaa057.896